|Apr 19, 2014, 01:33 AM|
Thank you for the inspiration
i saw your video and just had to copy the idea.
My bixler is powered by 2213 800 KV motors, with 9x3,8
multiwii with custom software.
custom mixer, with thr,ail,elev,rud,aux in - 2xail,2xflap,2xtilt,elev,rud servos and thr,ail,ele,rud,aux to multiwii.
arm/disarm are disabled in SF & FF, by using throttlecut & flightmodes in the TX
hover - all rudders are disablet and centered, flaps and aileron as far down as possible, control is 100% from multiwii
SF - all rudders double throw, flaps down, aileron slightly up, multiwii is dialed down to 50% control
FF - all rudders normal throw, flaps up, throttles are directly thru multiwii to controllers, gyros are disablet
i use only 5 ch from the TX
i made the motor hinges with adjustable endpoints, for positive horisontal and vertical control.
I had the exact problem yesterday, and broke pod and landing gear.
on the other side, the vtol bixler can do some IMPRESSIVE flatspinnning
of the four successfull transitions so far, two of them had indications that one servo was lagging behind. (severe yaw & roll to one side or the other)
in the video there are noticable yaw and roll to the left.
I used a D-MG16 Digital Metal Gear 2.9kg/.08sec/18.8gservo for tilt.
think i will cut them out, and replace with something more potent
|Apr 19, 2014, 09:05 PM|
Congratulations on getting your Bixler up and hovering. Several others have said they were building one, but you are the first to send video. I just this morning completed testing of my plane with the HS-225MG tilt servos, 66.65 oz. in. of stall torque at 6V, and I am providing them with 6V. They did the job for normal flying but I can still induce a failure if I dive while in hover or slow forward flight mode and build up a lot of airspeed before flipping the switch to transition into forward flight. For normal transitions from hover, or slow forward flight without diving first, they do fine. Make sure you have minimal drag in the linkages and tilt mechanism. If one side has more drag that the other it can cause a sharp turn and ultimately a crash. As you probably already know, pulling the throttle back to idle for even a moment relieves the pressure on the servo's and allows them to complete the transition.
Unfortunately, I do not understand 2.9kg as a measure of torque. Torque needs both a force and a lever arm to be meaningful and I don't know the length of the lever arm. I do know that 2.9kg = 102.3 oz. If I guess that it is 2.9 kg cm then that would be 40.27 oz. in. as compared to 48.6 oz. in. for the HS-85MG+ servos I used to begin with, and 66.65 oz. in. for the HS-225MG servos I am using now. By my measurements the Hitec specifications overstate the actual torque by about 30% and I have no idea of the actual torque of the D-MG16 servos you are using. It seems obvious that it isn't enough though.
Good luck, and don't dive into the outbound transition. This is one unusual case where too much airspeed is bad.
|Apr 20, 2014, 01:07 AM|
I saw your video only 1½ month ago
What i have also noticed is when i have a slow transitions with a stop at SF, it is more likely to happen - which meshes quite nicely with your observations.
If I hover and put it into a slight nosedown attitude, i can transitions without issues when i go directly from hover to FF with a 1-1½ sec transition.
But it doesn't look nice and controlled.
you and your imperial measures change to metric, like the rest of the world.
yes, a 2 kg servo, is 2 kg/cm - just another saying. Something like "to dremmel" meaning drilling or grinding something model related.
This bixler is only a steppingstone / development platform, for more computing power.
I am thinking about putting a pitot tube on it, to fully automate the transition based on speed input, basing the tilt on something wormdriven instead of geardriven,
before moving onto something with collective & cyclic control - i would LOVE to have something with more pitch in FF.
|Apr 21, 2014, 11:32 AM|
Also, I have my transitions set to 4 seconds in the TX, and they are very smooth.
Yes, we need to go metric. What's worse, we have a mix of both systems everywhere.
Glad to hear about your plans for more complex and interesting models. Consider a Separate Lift and Thrust (SLT) type as a much simpler way to get higher performance. You can use an optimized prop for forward thrust, and stop the lift props in line with the airflow for efficient cruse, all without the complexity of variable pitch.
Using airspeed as an input to control the transition is an idea we are toying with for the OpenAero2-VTOL firmware. It's probably not necessary for a simple sport VTOL, but the idea makes sense.
|Apr 21, 2014, 03:11 PM|
You might consider this as a linear drive option. They are not cheap, about $70.00 as I recall. Depending on the length of the stroke and the gear ratio they can be very slow, but wing tilt or rotor tilt is not something that needs to happen quickly. I did some analysis of the option that was the best fit for the VTOL-Trainer and while it had about 3X the force of the original servo I used, it also took about 3 seconds for full travel, both of which were acceptable, so it comes down to spending the money. They don't produce nearly as much mechanical power as a similar rotory servo, meaning speed x torque, but they can produce quite a lot of torque at low speed. They also hold strong when you try to backdrive them. All of which makes them pretty good for wing tilt, rotor tilt, and retractible landing gear type applications.
|Yesterday, 01:50 PM|
The eagletree airspeed sensor has i2c output
you just need to set it up to generic mode instead.
I use them for Multiplex Telemetry sensors.
I have looked at openaero, but not enough outputs, my bixler needs 12 channels out.
Maybe if it got a PCA9685 option in the SW, then i could have 8+16 channels output, but with 6 inputs, then it is almost too small.
My next idea needs 3xccpm, throttle, aileron and flaps in each wing + tilt, 2xelevator, 2xrudders, gear in the fuse, since i never use y-cables, i would need 18 outputs.
i am thinking about using a Cortex M3 for the next project.
i could surely use the 32bit / 72 MHz / 256K flash / 64 K Ram for that many channels - no more performance issues with kalman filters
besides, with y-cables it is impossible to do differential aileron throw or use the rudders as brakes.
Almost all my planes has differential throw on the ailerons, and my modified MPX Twinstar, has airbrakes mixed on the rudders.
Sorry to pull the thread away from VTOL bixlers......
|Category||Thread||Thread Starter||Forum||Replies||Last Post|
|Discussion||Another Success Story for the Obama Administration.||LcJ||Life, The Universe, and Politics||77||Jan 20, 2013 12:05 AM|
|Discussion||Another F35 trying to VTOL!||capt||VTOLs||13||Feb 07, 2010 12:59 PM|
|Discussion||EDF vtol..most successful||vortices||VTOLs||4||Mar 24, 2009 10:44 AM|
|Discussion||Yet another VTOL||Bobnormal||Foamies (Scratchbuilt)||29||Oct 27, 2008 05:50 AM|
|Vtol Plane Success||vtolman||Electric Plane Talk||34||Jan 11, 2006 01:01 PM|